Article Text

Download PDFPDF

Auscultation in the diagnosis of respiratory disease in the 21st century
  1. Claudia C Ceresa1,
  2. Ian D A Johnston2
  1. 1
    Division of Therapeutics and Molecular Medicine, University of Nottingham, Queen’s Medical Centre, Nottingham, UK
  2. 2
    Department of Respiratory Medicine, Nottingham University Hospitals NHS Trust, Queen’s Medical Centre Campus, Nottingham, UK
  1. Dr I Johnston, Department of Respiratory Medicine, Nottingham University Hospitals NHS Trust, Queen’s Medical Centre Campus, Nottingham NG7 2UH, UK; ian.johnston{at}

Statistics from

Auscultation of the chest via a stethoscope has been an integral part of respiratory examination for the last 200 years. Hippocrates was known to “directly auscultate” the chest by placing an ear to the patient’s chest wall, but this had fallen out of favour until the 1800s, when Laënnec invented the precursor for the modern day stethoscope in 1816. This consisted of a wooden rod, one end of which was placed on to the chest wall and the other to the doctor’s ear. In 1819, Laënnec went on to describe lung sounds and compare them with pathology found at autopsy. He described five types of chest “rattles” (rales), but, in part because of translation problems, there was subsequent confusion around what was meant by dry or wet crepitations and rhonchi.1 Fortunately, owing in large part to the seminal work of Forgacs,2 the sounds heard at auscultation were simplified into breath sounds (normal or bronchial), adventitious sounds (crackles, wheezes and rhonchi) and voice sounds (normal and bronchophony). Various other sounds such as pleural rubs, clicks and squawks may also be heard.2 3


Even today the breath sounds are incompletely understood, but the normal breath sounds are probably the result of turbulent airflow through the larger airways.2 3 At the chest wall, the higher frequencies are attenuated by the lung parenchyma, but, in the presence of consolidation, these higher frequencies are transmitted, leading to “bronchial breathing”. The same mechanism leads to bronchophony in which the spoken word is heard clearly because of this transmission of higher frequency sounds. Whispering pectoriloquy is a redundant sign, as it occurs via the same mechanism and adds no further information. By contrast, the breath sounds are reduced or absent when the lung and chest wall are separated by fluid or air, as in effusion or pneumothorax.2

Wheeze is a musical sound with duration longer than 250 ms and is probably due to airway flutter caused by airway narrowing. Some still refer to rhonchi, which are low-pitched musical sounds, probably caused by airway secretions.1 4

Crackles are intermittent explosive sounds not lasting longer than 20 ms.3 Although initially thought to be due to air bubbling through secretions, it is now appreciated that crackles are mainly the result of closed airways opening suddenly during inspiration or closing during expiration to equilibrate pressures within the airway.2 3 This occurs as a result of increased elastic recoil, inflammation or oedema, and thus the characteristics of the sound produced may depend on the diameter of the airway.5


The advantage of evaluating lung sounds with a stethoscope is that it is relatively cheap and non-invasive and it is expected by the patient. The lung sounds found in various pathologies are highlighted in major texts,4 and teaching auscultation remains an absolutely standard feature of medical education. It is all the more surprising therefore to find a dearth of research as to the diagnostic value of the lung sounds, pneumonia and lung fibrosis being perhaps the most common conditions studied. In pneumonia, Lieberman et al6 reported that clinical assessment had a positive predictive value of only 27% in diagnosing pneumonia. Others have also noted that chest examination is a poor diagnostic tool in pneumonia,7 8 a chest radiograph still being required to confirm or refute the diagnosis.7 Rates of clinical diagnosis of pneumonia have run between 7% and 38%.9 Surprisingly, however, as chronicled in this edition of the Journal, (see page 432) late inspiratory crackles were recently found to have a positive predictive value of 76% in differentiating “bacterial” from “atypical” pneumonia.10 Inspiratory crackles occur in over 80% of patients with idiopathic pulmonary fibrosis, but feature as only a minor diagnostic criterion.11 Wheeze usually indicates airflow obstruction, but may be absent, particularly in severe disease.12 The stethoscope can obviously provide useful information in certain contexts. For example, findings such as wheeze or crackles may often point to appropriate investigation or treatment of the breathless patient. However, its diagnostic usefulness is clearly limited, and highly dependent on the user, with considerable observer variability.13


Given this background, investigators have developed objective methods of recording and analysing lung sounds, to both understand their origin and to enhance their diagnostic utility. Such techniques began with simple microphones and tape recorders in the 1960s and progressed to digital signal-processing techniques3 14 Computer-assisted mapping of lung sounds continues to develop, with, for example, recent work on vibration response imaging, which displays lung sound distribution as a dynamic image.15 16 Although the results are objective, there are difficulties comparing studies because of differing techniques in different laboratories, as often the equipment used is developed in the laboratory itself or equipment designed for other uses is modified.17

These studies have greatly increased our knowledge of the origin and nature of lung sounds, but what has been the clinical impact? Crackles in fibrosing alveolitis or asbestosis can, for example, be distinguished from those in conditions such as bronchiectasis, chronic obstructive pulmonary disease and heart failure.18 19 These and other small studies in the clinical arena have consistently concluded that non-stethoscopic evaluation may be of diagnostic value. However, there are no large-scale clinical investigations, possibly in part as a result of a lack of commercial interest. Until studies are available that quantify the diagnostic value in less selected populations, the clinical impact of non-stethoscopic recordings of lung sounds will remain, as now, minimal.


Given this rather negative appraisal of the value of lung sounds, and the easy access to imaging and lung function, it might be concluded that auscultation should be abandoned. However, this would be a mistake—investigations are not always to hand, particularly in primary care and the developing world. Auscultation is very cheap and can provide valuable information as part of an overall assessment and even on its own, provided that its limitations are understood.

Auscultation in the 21st century is little changed from the last century, although the terminology is simpler. It should continue to be taught and practised. Sophisticated methods of recording lung sounds are unlikely to enter the clinical arena in the near future.


View Abstract


  • Funding: CC is funded by Asthma UK.

  • Competing interests: None.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles